---

Major lower extremity amputations related to peripheral arterial disease and diabetes mellitus in the city of Rio de Janeiro
(Portuguese PDF version)

David Spichler¹, Fausto Miranda Jr.², Ethel Stambovsky Spichler³, Laércio Joel Franco⁴

1. MD. PhD in Vascular Surgery, Escola Paulista de Medicina, Universidade Federal de São Paulo (EPM/UNIFESP), São Paulo, SP, Brazil.
2. MD. Head professor of Vascular Surgery, EPM/UNIFESP, São Paulo, SP, Brazil.
3. MD. PhD in Endocrinology, EPM/UNIFESP, São Paulo, SP, Brazil.
4. MD. Professor of Preventive Medicine, Faculdade de Medicina de Ribeirão Preto, Universidade de São Paulo (USP), Ribeirão Preto, SP, Brazil.

Financial support: Ministério da Saúde/Escola Paulista de Medicina - UNIFESP.

Institutional support: Ministry of Health, Health State Secretariat of Rio de Janeiro, DataSUS, Brazilian Diabetes Society, Escola Paulista de Medicina - UNIFESP, University of Pittsburgh, PA, USA - Pan American Health Organization (PAHO).

Correspondence:
David Spichler
Rua Barão de Icaraí, 33/1306
CEP 22250-110 - Rio de Janeiro, RJ, Brazil
Phone: +55 (21) 2552.1012/2551.5791
E-mail: spichler@terra.com.br

---

ABSTRACT

Objective: To estimate incidence and levels of 4,818 major lower extremity amputations related to peripheral arterial disease and diabetes mellitus, performed in the city of Rio de Janeiro between the years of 1990 and 2000.

Methods: Analysis of amputee registry and medical records, considering the following data: age, gender, amputation level, incidence rate from passive surveillance, and capture-recapture systems. The Chi-square ( Χ²) test was used to compare proportions.

Results: Primary major lower extremity amputations were 97.7% and the secondary ones 2.3%. In 43 hospitals included in the study, 56.3% of amputations were caused by peripheral arterial disease whereas 43.7% were due to diabetes mellitus (P < 0.001), with frequency increasing from 1.2% to 22.9% for diabetes mellitus and from 4% to 19.4% for peripheral arterial diseases (P < 0,001). Males (M) and age groups ranging from 65-69 and 55-79 years influenced peripheral arterial disease (P < 0,001). The average age was 64.89 (± 10.35) years for diabetes mellitus, and 66.36 (± 11.90) years for peripheral arterial diseases (P < 0.001). Primary above-knee major lower extremity amputations were 71.8%, with 59.9% caused by peripheral arterial diseases and 40.1% by diabetes mellitus (P < 0,001). The ratio above/bellow-knee was 3.2:1 (peripheral arterial diseases), and 1.9:1 (diabetes mellitus). Incidence rates increased five times; 18 and 3 times for diabetes mellitus and 19 and 2.2 times for peripheral arterial disease, in males and females respectively. Capture-recapture results were 20%, 370% and 350% higher (55-74 years) for the general population, diabetes mellitus, and peripheral arterial disease.

Conclusion: The increased number of amputations and the high level of primary major lower extremity amputations, which represent five times increase in frequency of this type of amputation, should be considered an important public health problem.

Key-words: amputation, lower extremity, incidence, diabetes mellitus.
Palavras-chave: amputação, extremidades inferiores, incidência, diabetes melito.

J Vasc Br 2004;3(2):111-22

---

Major lower extremity amputations (MLEA) represent a significant socioeconomic impact, involving loss of labor capacity, sociability and consequently, decrease of quality of living. They are among of the most devastating complications of degenerative chronic disease, associated with high morbidity, incapacity and mortality.

Atherosclerotic vascular disease, which affects the lower limbs, is the most common manifestation of peripheral arterial disease (PAD), causing clinical conditions that may vary from intermittent claudication or rest pain to ulceration and gangrene.¹

Peripheral vascular disease from the International Classification of Diseases (ICD) 9 codes 440-59 are much studied in industrialized countries, whereas in developing countries, researchers give preference to other types of cardiovascular diseases. In Brazil, for instance, there is no reference to any epidemiological study about PAD² and literature review is scarce when critical ischemia is not associated with diabetes mellitus (DM).

The aim of this study was to analyze the characteristics of MLEA, their frequency, some demographic variables and amputation levels. The MLEA analyzed in this study were performed in the city of Rio de Janeiro from 1990 to 2000.

PATIENTS AND METHODS

All MLEA performed from January 1st, 1990 to December 31st, 2000 were retrospectively reviewed with the purpose of assessing clinical variables.

The data about MLEA were obtained from the partial burying statement, hereafter called amputation registry (AR), issued by the Health State Secretariat of Rio de Janeiro.

The source of information was 43 hospitals. Twenty-one public health institutions belonging to the Health System (municipal, state and federal) provided 65.9% of MLEA, four institutions from the Armed Forces contributed with 5.7%, two university institutions with 10.4% and 16 private institutions provided information on 17.9% of amputations.

The demographic variables, such as gender identification, age, color, and residence are described on the AR.

Color was defined as white and non-white. However, this information was interrupted from the second semester of 1996 onwards, since the item color was excluded from the AR.

Amputation etiology was classified according to the ICD 9 and 10. The ICD-9 codes 4402 and 7854, and ICD-10 codes I700-3 were used to represent PAD. ICD-9 codes 2500, 7850, 4438, and ICD-10 codes E115-118, E145, I738-9 were used to represent DM. For analysis purposes, ICD-10 codes were converted to ICD-9.

The terminology employed to describe the three sources of research used in this study was peripheral arterial disease (PAD), symptomatic or critical ischemia, and absence of any reference to diabetes mellitus. We analyzed lower limb amputations caused by diabetes mellitus and excluded any other causes. We considered DM diagnosis the diabetic patient who presented with critical ischemia as the basic cause. Diabetes, gangrene, hyperglycemia, or any reference to DM were considered secondary cause.¹

The study, designed to obtain an estimate of incidence, was based on the number of registered inhabitants of 8,543,188 and 8,682,474 defined by the Brazilian Institute of Geography and Statistics (IBGE) for the second half of the years 1993 and 1996, respectively.

Age group 30-89 years was estimated as of 4,442,457 and 4,514,886 inhabitants. Age group 55-74 years was estimated as of 905,577 e 920,342 inhabitants, both for the years 1993 and 1996, respectively. This data served as a basis for prevalence calculation of DM and PAD, respectively. The division into these two age groups is justified by previous population-based studies of prevalence, claudication, and critical ischemia due to PAD. Focus was put on 55-74 age group for comparative effect of our patients with studies developed by different authors.^1,3

In order to estimate the diabetic population by age group, we used information from the diabetes mellitus census, conducted in Rio de Janeiro,⁴ taken in consideration the self-referred prevalence of 9.2% for the population between 30-89 years of age. For age group 55-74 years old, the diabetic and the PAD population represent 15% and 17% of the population group in this age group, respectively.¹

The population with claudication, used as estimate in cases of PAD associated with DM, was evaluated in 23.55% of the population belonging to age group 55-74 years.¹

Incidence estimate of amputations caused by PAD and DM was conducted with the use of the capture-recapture technique (C-R).⁵ It consisted of a retrospective evaluation of MLEA performed between January 1st, 1992 and December 31st, 1994 in the city of Rio de Janeiro. Three sources were employed: major source or source 1, with 1,191 ARs from 23 hospitals; source 2, from 157 medical records provided by the prosthesis center (Centro de Referência de Protetização da Associação Brasileira de Reabilitação do Rio de Janeiro - ABBR); and source 3, from 34 medical records provided by the rehabilitation center (Centro de Reabilitação e Fisioterapia Municipal Instituto Oscar Clark).

Definition of amputation

Only the amputations defined as "major' were analyzed, in other words, the amputations of the proximal portion of the foot, leg (below-the-knee amputation), thigh (above-the-knee amputation) and hip disarticulation.

The side (left or right) and the level of amputation (proximal portion of the foot, leg, thigh, or hip disarticulation) were recorded in chronological order. The amputations are referred as primary, secondary, and uni or bilateral.

Primary amputation is defined as an amputation of a segment of the ischemic lower limb, without any antecedents of amputation procedure or revascularization. Secondary amputation is defined by the presence of any previous procedure like thrombolysis, angioplasty, revascularization, or when another amputation of the same limb occur.³

Statistical analysis

The GLIM⁶ software has a statistical analysis interactive model, developed by the Royal Statistical Society. It is used to analyze log-linear logistic regression models for contingency and survival tables. This software was employed in the capture-recapture technique to analyze the three sources of information.

In order to analyze the variables gender, color, age group and other variables related to MLEA, the Χ² (chi-square) test was used with a significance of P < 0.05.

Data Bank

The information obtained was processed using the EPI INFO 6.4 software. In order to control errors in this process, double typing was performed and the routines provided by EPI INFO 6.4 were used.

RESULTS

During the time period under study (1990-2000), 5,539 major MLEA were performed due to six etiologies: 2,853 cases of PAD (51.5%), 2,170 cases of DM (39,2%), 314 traumas (5.6%), 93 cases of osteomyelitis (1.7%), 66 cases of gas or emphysematous gangrene (1.2%), and 43 cases of neoplasia or tumor (0.8%). Among 5,023 (90,1%) MLEA caused by PAD and DM performed in 4,878 patients, 205 amputations were excluded (4.1%) based on the following reasons: 93 (1.9%) because the age of the patient was not specified on the form; 21 (0.4%) patients were under 30 years old (16 related to PAD and 5 related to DM), and 91 (1.8%) patients were over 89 years old (63 related to PAD and 28 related to DM).

Thus, 4,673 (95.8%) patients were enrolled, being 2,631 (56.3%) PAD and 2,042 (43.7%) DM. Among those, 145 (3%) presented primary bilateral amputations, totaling 4,818 MLEA. Out of these 4,818 MLEA, 4,707 (97.7%) are primary amputations and 111 (2.3%) are secondary, caused by PAD (56.3%) and DM (43.7%), for age group from 30 to 89 years old.

Table 1 shows the total number of amputations due to PAD and DM per year. Historical series presents a higher number of amputations caused by a significant increase of reporting units. In 1997 and 1999 occurred an under-reporting of the number of hospitals and amputations.

Table 1 - City of Rio de Janeiro. Time period 1990-2000: Frequency of MLEA due to DM e PAD (30-89 years)

Period	DM		PAD		Total
	n	(%)	n	(%)	n	(%)
1990	27	1.3	167	6.1	194	4.1
1991	43	2.0	181	6.7	224	4.7
1992	34	1.6	228	8.4	262	5.4
1993	58	2.8	260	9.6	318	6.6
1994	236	11.2	289	10.6	525	10.9
1995	227	10.8	201	7.4	428	8.9
1996	304	14.5	227	8.3	531	11.0
1997	90	4.3	81	3.0	171	3.5
1998	420	20.0	396	14.6	816	16.9
1999	179	8.5	233	8.6	412	8.6
2000	483	23.0	454	16.7	937	19.4
Total	2,101	100	2,717	100	4,818	100

1997 and 1999 under-reported; PDD > DM (P < 0,001); PAD 1990-1994 (P < 0,001); DM 1995-2000 (P < 0,001).

The average age of the 4,818 patients was 65.73 (± 11.28) years. The average age of males was 64 (± 11) years, and that of females was 67 (± 11) years (P < 0,001). For diabetics, the average age was 64.89 (± 10.35) years and for PAD patients 66.36 (± 11.90) years (P < 0,001).

Regarding gender, 2,788 (57.9%) patients were males. As for the level of amputation due to PAD or DM, 3,460 (71.8%) were on the thigh, 1,137 (23.6%) on the leg, and 219 on the foot. Two disarticulations occurred.

For age distribution, the ratio male/female was 1.4(P < 0,001).

With respect to age group, the MLEA due to PAD and DM were more frequent between 65-69 years, with 910 (18.9%) amputations, followed by age group 60-64 years with 765 (15.9%), and 55-59 and 70-74 years with 654 and 658 amputations, respectively, representing 13.5% of the total number of amputation due to PAD and DM. In age group 55-74 years occurred 62% of the MLEA.

Figure 1 shows the amputation levels according to the etiology (PAD and DM) for the time period of 1990-2000.

Figure 1 - Ratio of 4.818 MLEA (30-89 years) according to levels and etiology 1990-2000

As a primary procedure, the MLEA due to PAD and DM presented a frequency of 70% for thigh, 23.1% for leg and 4.5% for foot.

Regarding MLEA on the thigh, there was a predominance of MLEA due to PAD (2,070 or 59.9%) over those due to DM (1,390 or 40.1%), with a 1.5:1 ratio (P < 0,001).

The ratio of above/below-knee MLEA due to PAD and DM is 2.5:1 (P < 0,001). Among diabetics, this rate is 2:1, and in cases of PAD, the rate is 3.2:1. Among 4,818 MLEA due to PAD and DM, 145 (3%) were bilateral with primary procedure whereas 111 (2.3%) were secondary procedures.

Male predominance is maintained among secondary MLEA, with 61.6%, and bilateral with 55.2%.

The skin color of the patients was analyzed from January 1st, 1990 to June 1st, 1994 in 1,185 (82.3%) MLEA. The color was referred as white in 680 (57.4%) and non-white in 505 (42.6%) amputations. Ratio white/non-white for this time period was 1.3:1 (P < 0,001).

Regarding the side of amputation, 2,339 (48.5%) were performed on the left side and 2,479 on the right side, without significant statistical difference.

Table 2 shows yearly incidence per 100,000 inhabitants of MLEA caused by PAD and DM.

Table 2 - Annual incidence of MLEA due to DM and PAD in 100, 000 inhabitants in age groups between 30 and 89 and 55 and 74 years in the city of Rio de Janeiro from1990 to 2000

Year	DM/100,000 30-89 years*	PAD/100,000 30-89 years*	DM/100,000 55-74 years*	PAD/100,000 55-74 years*
1990	0.60 (0.55-0.66)	3.70 (3.58-3.81)	1.74 (1.66-1.82)	11.86 (11.66-12.06)
1991	0.95 (0.89-1.01)	4.00 (3.87-4.12)	3.04 (2.94-3.14)	12.49 (11.80-13.18)
1992	0.75 (0.70-0.80)	5.05 (4.91-5.18)	2.39 (2.30-2.48)	15.32 (14.56-16.08)
1993	1.28 (1.21-1.34)	5.76 (5.61-5.90)	3.69 (3.57-3.81)	17.60 (17.34-17.86)
1994	5.22 (5.08-5.36)	6.40 (6.24-6.55)	18.58 (18.32-18.84)	19.12 (18.75-19.37)
1995	5.03 (4.89-5.16)	4.45 (4.32-4.58)	16.62 (16.37-16.87)	14.34 (13.60-15.08)
1996	6.73 (6.57-6.88)	5.03 (4.89-5.16)	31.73 (31.39-32.07)	12.71 (12.49-12.93)
1997	1.99 (1.90-2.97)	1.79 (1.70-1.87)	5.76 (5.62-5.90)	4.45 (4.32-4.58)
1998	9.30 (9.11-9.48)	8.77 (8.58-8.95)	31.55 (31.16-31.86)	25.09 (24.78-25.40)
1999	3.96 (3.83-4.08)	5.16 (5.02-5.30)	12.60 (12.38-12.82)	13.79 (13.56-14.02)
2000	10.70 (10.49-10.90)	10,05 (9,85-10,24)	33.14 (32.08-34.36)	26.62 (26.30-26.94)

* Confidence interval (CI) 95%.
1997 and 1999 under-reporting; population between 30 and 89 years = 4,514,886; PAD 55-74 years = 10,6% of the annual population (1990-1994); PAD > DM (55-74 years), (P < 0,001)
(1990-1994); DM > PAD (55-74 years), (P < 0,001) (1995-2000).

The incidence of both DM and PAD MLEA shows an increase. This trend, at the end of the study period, is respectively 17.8 and 2.7 times higher for age group 30-89 years (P < 0,001), and 19 and 2.2 times higher for age group 55-74 years (P < 0,001).

In addition, it was observed that from 1995 onwards, for age group 55-74 years, the incidence of MLEA due to DM exceeded the incidence of MLEA caused by PAD (P < 0,001).

Table 3 shows incidence estimate of MLEA for the general population and diabetic population in the age group 30-89 years. Also, for the general and PAD population, in age group 55-74, during 1990-2000.

Table 3 - Incidence of MLEA per 100,000 inhabitants a year according to age groups, diabetes and peripheral arterial disease in the city of Rio de Janeiro

Age group	Period	General Population*	Diabetic population*	PAD population*	DM+PAD population*
30-89	1992-1994	8.29 (8.12-8.46)	26.75 (26.36-26.98)	(*)	(*)
30-89	1990-2000	9.70 (9.51-9.89)	45.98 (45.57-46.39)	(*)	(*)
55-74	1992-1994	25.98 (25.69-26.29)	55.69 (55.23-56.15)	103.71 (103.09-104.33)	110.58 (109.94-111.22)
55-74	1990-2000	29.50 (29.17-29.83)	92.19 (91.60-92.78)	92.21 (91.62-92.80)	170,30 (169.50-171.10)

(*) Without population-base registry.
* Confidence interval of 95%.
30-89 years DM > general population, (P < 0,001); 55-74 years PAD > DM, (P < 0,001); 55-74 years DM associated with PAD > PAD, (P < 0,001).

For age group 30-89, the incidence is 4.7 times higher among the diabetic population than in the general population, during the time period 1990-2000 (P < 0,001).

In the same period, for age group 55-74 years, the incidence rates in diabetics and in patients with PAD are three times higher than the incidence rate among the general population (P < 0,001).

For DM and PAD associated, the incidence for age group 55-74, during the period 1990-2000, was six times higher than the incidence for the general population, and two times higher than the incidence for DM and PAD.

MLEA incidence among males was higher in all related periods (P < 0,001).

The incidence of MLEA due to PAD, during period 1992-1994, showed a male/female ratio of 2:1, and during 1990-2000 the ratio was 1.7:1.

Table 4 shows the results of frequency found by C-R. When compared by passive surveillance for age group 30-89 years, related to the general population, DM, and PAD, the results were 3.7, 3.8, and 3.7 times higher, respectively. For age group 55-74 years, in relation to the general population, DM, and PAD, the results obtained were 3.5, 3.7, and 3.4 times higher, respectively.

Table 4 - Number of MLEA found through passive surveillance and estimated by Capture-Recapture technique, for age groups 30-89 and 55-74 years, time period from 1992 to 1994 in the city of Rio de Janeiro

Etiology	PS	C-R	SD	Estimated C-R	IC 95%
30-89 years	1,105	4,108	6.72	4,125	4,100-4,152
PAD	777	2,858	6.02	2,872	2,865-2,881
DM	328	1,250	1.16	1,253	1,252-1,257
55-74 years	706	2,500	4.97	2,505	2,490-2,510
PAD	479	1,650	0.52	1,654	1,651-1,661
DM	227	850	4.41	852	850-855

PS = Passive surveillance; C-R = Capture -Recapture; SD = Standard deviation; CI = Confidence interval of 95%.

Table 5 shows a comparison between incidence estimates, per 100,000 inhabitants a year, during 1992-1994, through passive surveillance and capture-recapture technique (C-R).

Table 5 - Comparative incidence of MLEA through C-R and PS according to etiology for age groups 30-89 years and 55-74 years in the city of Rio de Janeiro between 1992 e 1994

	Estimated population	C-R/100,000/year*	PS/100,000/year*
General population (30-89)	4,442,457	9.40 (9.21-9.59)	8,29 (8,12-8,46)
DM	408,706†	102.19 (101.57-102.81)	26.75 (26.36-26.98)
PAD	4,442,457	21.54 (21.26-21.82)	(-)
General population (55-74)	905,577	31.36 (31.02-31.70)	25.98 (25.69-26.29)
DM	135,836‡	209.07 (208.18-209.96)	55.69 (55.23-56.15)
PAD	153,948§	358.99 (356.96-359.28)	103.71 (103.09-104.33)

* Confidence interval of 95%.
†Calculation of 9.2% of population between 30-89 years, from the second half of 1993.
‡Calculation of 15% of population between 55 and 74 years.
§Calculation of 17% of population between 55 and 74 years.
(-) Without population based registry.

The C-R technique revealed that, for age group 55-74, the incidence was 6.7 and 11.4 times higher in the diabetic and PAD population, respectively.

The passive surveillance showed that, for age group 30-89 years, the incidence was 3.2 times higher in the diabetic population when compared to the general population, whereas the C-R technique showed an incidence 10.9 times higher

DISCUSSION

Clinical evolution

One of the major factors related to the clinical evolution to MLEA is the impropriety of an early clinical diagnosis of PAD and diabetes mellitus

Hirsch et al.⁷ evaluated 6,979 patients based on their medical history and their ankle-brachial pressure index (ABI). The patients' age groups were 70 years or over, between 50 and 69, and smoking and diabetes mellitus history. The authors show higher prevalence of PAD in the primary clinical practice, although the majority of the physicians do no take the diagnosis of PAD into consideration. A simple checking of the ankle-brachial pressure index was enough to identify a great number of patients with PAD who had not been previously identified. Hirsch et al. emphasized that the data suggest that the physicians who exclusively use a classical history of claudication to detect PAD are unable to identify 85 to 90% of PAD cases. Paradoxically, there is a high diagnostic difficulty in, at least, 50% of patients treated in the primary care network. This fact causes an under-reporting of intermittent claudication, as well as critical ischemia and its treatment, reflecting an existing barrier within the secondary and tertiary care and a risk increase of ischemic events, amputations, and death.

McLafferty et al.,⁸ in a study developed in the Southwestern Illinois in the United States, observed that patients with PAD are assigned later to surgical treatment due to the lack of an efficient early diagnosis. This fact has a disadvantageous effect on the results. The authors also discussed that, after a test with clinical internists, they found that only 37%, that is, one third of patients had a diagnosis of PAD, and 26% had a diagnosis of ulcerating wound on the foot, compared to 90% of diagnosis of cardiovascular and pulmonary diseases, diabetes mellitus and cerebrovascular accident. Only 34% of these patients had their abdominal aorta and their distal peripheral pulses examined in the absence of peripheral pulses. Doppler ultrasound and ABI were conducted in 25% of these patients.

In a comparative table, McLafferty et al.⁸ showed that after diagnosis of PAD, 9% of carotid arterial disease and aneurysm of the abdominal aorta and 14% of critical ischemia in PAD were diagnosed. Half of these cases were referred to vascular surgical treatment, 33% were referred to the general surgeon, 13% to the cardiothoracic surgeon, and 1% was referred to the radiologist, indicating a the need of better training for clinicians, residents and medical students regarding diagnosis and early referral of these patients.

In another study developed in Chicago, in the Northwestern University of Illinois, Ebaugh et al.⁹ analyzed 16,422 patients for a period of seven years (1993-1999). The patients had undergone aortoiliofemoral bypass, femoro-popliteal bypass, or distal revascularization, and major lower-limb amputations after surgical procedure. The authors evaluated which hospitals with high qualification for performing complex vascular procedures present lower rates of postoperative complications, including reduction of major lower-limb amputations. The hospitals that offered cardiac surgery, vascular flow laboratory, general surgery residence or training in vascular surgery were considered as highly capable of performing a complex vascular procedure. The authors also selected 16 out of 98 hospitals, which means 34% of the procedures performed, including eight hospitals with more than 40 distal revascularization procedures per year. Hospitals were classified as presenting low number versus high number of complications; mortality rate of 2.8% versus 3.8% (P = 0,003); amputation rate of 4.6% versus 4.9% (without statistical significance), but high number of complications, with 9.8% versus 8.5% (P = 0,006). They concluded that mortality was higher in hospitals with higher capacity of solving more complex cases, since a higher number of patients with more complex clinical conditions are treated. They also conclude that mortality was higher in those hospitals where the number of complications does no present any correlation with mortality, which means that such hospitals do not provide a reliable qualified care.

Another important factor that affects ad therapeutics and results was described by Connelly et al.¹⁰ from the University of Leeds in the United Kingdom. The authors selected six cases, being two cases with indication for primary MLEA, two with indication for revascularization and 50% probability of lower-limb preservation, and two cases with 80% probability of lower-limb preservation. The patients were submitted to 10 vascular surgeons from different places in England. The results regarding diagnosis, management, clinical evolution and prognosis were in agreement in 40% of cases, non-conform in 40%, and totally diverse in 20%. These oscillating results from MLEA showed that treatment varies according to the experience of the vascular surgeon and the location where he/she works.

Regarding the experience and conducts of the vascular surgeon in the evaluation of patients with lower-limb critical ischemia, the Vascular Surgical Society of Great Britain and Ireland¹¹ developed a study that divided vascular surgeons into four groups according to the annual number of revascularization procedures performed: 0-10, group I; 11-20, group II; 21-30, group III; and over 30, group IV. Vascular surgeons participated in the retrospective study for a period of three months. Defined basic criterion was the need of surgical procedure in order to avoid or postpone lower-limb amputation. Revascularization procedures included: percutaneous transluminal angioplasty, the use of thrombolytics, and femoro-popliteal distal bypasses. It was concluded that surgeons from group I, with less than 10 vascular procedures per year, tend to present a lower number of revascularization procedures and a higher number of primary amputations than surgeons from other groups. However, it was emphasized that the number of indications for angioplasty was similar in the four groups.

Cosgrove et al.¹² also developed a study on the experience of vascular surgeons. The authors observed that MLEA performed by younger surgeons required repeated re-examining and re-amputations at levels above the primary MLEA, when compared to MLEA performed by experienced surgeons.

Around 60 to 90% of patients presented with critical ischemia undergo some kind of revascularization procedure.¹³ Among these patients, 25% were diabetics.¹⁴ They presented with seven times more MLEA than patients with PAD.¹⁵

In a study developed in the city of Rio de Janeiro, it was observed that 14% of patients with critical ischemia underwent a revascularization procedure such as femoro-popliteal bypass. Among these procedures, only 4% were performed in diabetics, who presented 10 times more MLEA than patients with PAD.¹⁶

According to Widmer et al.,¹⁷ patients with intermittent claudication presented MLEA around 1.8%; according to Peabody et al.,¹⁸ 2.5%, and according to Weitz et al.,¹ 4%, although, approximately 10-40% of patients with critical ischemia undergo primary MLEA and a small number receives clinical treatment.

Many studies confirmed the fact that patients who are initially classified as intermittent claudication were classified as critical ischemia after 1 or 2 years of follow up.^1,11,19

Wolfe,²⁰ in the United Kingdom, developed a one-year study with patients presented with critical ischemia. He noticed a 20% mortality rate among these patients.

Similar results were presented in a six-month study developed in Zurich.²¹

In China, Cheng et al.,²² in a prospective study involving 665 patients over 70 years of age with lower-limb critical ischemia, observed mortality rates of 15%, 20% and 45% with 1, 3 and 5 years, respectively. The authors emphasized age factor, ABI < 0.5, diabetes mellitus, and cardiovascular and renal disease as major risk factors that influenced mortality.

Another prospective multicentric study, conducted in Italy, showed that, after a 3-month observation period of patients with critical ischemia, 8.7% died, 12.2% underwent MLEA, whereas 17.9% remained with critical ischemia.

Level of MLEA

A review of the last 30 years has showed that the ratio of MLEA above-knee/below-knee is approximately 1, remaining unaltered for the past decades.³

Toursarkissian et al.,²⁴ in a retrospective study, with 56% of primary MLEA in 99 males, related ratio was 3:2.

Among our patients, this ratio is higher (2.5:1), probably due to the primary indication for thigh MLEA around 70%.

Faries et al.²⁵ showed that the employment of aggressive vascular procedures may alter these levels, with a decrease of 2.1 to 0.14.

Peel & Stonebridge,²⁶ in Scotland, concluded that above-knee amputations are more frequently performed in elderly patients, being more associated with high mortality and reduced survival.

In a study conducted by Mayfield et al.,²⁷ the authors analyzed 70,200 amputations, being 72% performed on the thigh and 52%of these amputations in patients with PAD, in contrast with 19.9% in diabetic patients. However, leg amputations were performed in 24.8% and 28.6% in PAD and diabetic patients, respectively. Foot amputations were performed in 4.7% in patients with PAD and 10.6% in diabetics. They also emphasized that minor amputations (toes and distal portion of the foot) were performed in 18.6% of patients with PAD, and 41.8% in diabetics.

Among our patients, similar results were found. MLEA performed on the thigh represent 71% of all MLEA due to PAD and DM, among which 43% were due to PAD and 28% due to DM. Leg amputation represented 23% of the MLEA, with 12% and 11% due to DM and PAD, respectively. On the foot, 4.5% of the MLEA, being 2.4% and 2.1% due to DM and PAD, respectively.

MLEA performed on the thigh has been preferentially used as primary indication. In our study, the results related to etiology and level of MLEA were similar to those referred by Mayfield et al.,²⁷ and others.^22,28

Faries et al.²⁵ showed that primary healing of above-knee MLEA occurred between 30-92% of amputations, with an average of 70-75%. One re-amputation occurred between 4 and 30% of patients

Approximately 15% developed secondary healing, requiring debridement and other procedures to afford a viable stump.²⁷

In 30% of MLEA performed below-knee and did not present primary healing, approximately half required another above-knee amputation.²⁹

Kihn et al.³⁰ noticed that in 4% of the cases which occurred primary healing of the below-knee MLEA, a re-amputation had to be performed on a higher level.

A studied developed in 51 hospitals of six European countries analyzed 713 patients with below-knee MLEA. After 3 months, 59% of patients presented primary healed stump, 19% needed re-amputation above the level of the MLEA on the leg or above-knee, and 11% remained with open unhealed wound.

Re-amputation was necessary in 2.3% of our patients, being more frequent on the thigh (65%) and in males 61.6%. These findings are probably due to the preference that is given to MLEA, primarily on the thigh, being 23% on the leg and 4.5% on the foot.

Among MLEA, 3% were bilaterally performed as primary indication, being 83% on the thigh and 55.2% in males.

The increased number of primary below-knee MLEA, probably related to complications, requires a higher number of amputations and, therefore, a higher number of thigh MLEA.

Prognosis and progression of the amputee

The array of inappropriate conducts that lead to avoidable major amputations was observed in 70% of diabetics in the city of Rio de Janeiro.³²

Therefore, there is a real perspective of prevention through the intervention in factors such as inadequate metabolic control, dyslipidemia, tabagism and education of patients with "foot risk".³

Based on these evidences, one of the preventive policies is a 50% decrease in amputations in individuals with peripheral arterial disease and diabetes. This is achieved through researches and programs, specialized multidisciplinary and interdisciplinary team training, as well as the implementation of "foot risk" ambulatories. Cost-benefit analysis suggest that preventive policies mean a much lower onus than the financial and social impact brought by disability or early death.³²

According to Kihn et al.³⁰ observations during the past 20 years, patients with below-knee MLEA tend to present faster rehabilitation than those who underwent above-knee amputations.

However, in our study performed in the city of Rio de Janeiro, MLEA were primarily performed on the thigh, remaining stable for 11 years, probably due to the critical condition of the patients at admission.

A great number of patients present economic and social problems, in addition to a mortality rate of 50% observed in a period of 3 years, in the city of Rio de Janeiro.³²

Kihn et al.³⁰ found that only 25% of amputees present distal pulses (pedis and posterior tibial) on the contra-lateral limb, and approximately 15% of patients will need a MLEA on the contra-lateral limb in a two year period.

Incidence and prevalence of critical ischemia and MLEA

The incidence of lower-limb critical ischemia can be estimated by the numbers of MLEA performed. According to Weitz et al.,¹ assuming that 90% of all MLEA are performed due to critical ischemia and 25% of patients with intermittent claudication will need a MLEA, we estimate that the incidence of critical ischemia is approximately 500 to 1,000/1,000,000/year.

Based on these parameters, we calculated the incidence of critical ischemia of 400/1,000,000/year in the population of Rio de Janeiro, with results similar to Denmark,³³ United Kingdon, and Ireland.³⁴

In one year, one out of 100 patients with intermittent claudication will develop lower-limb critical ischemia. The prevalence of intermittent claudication is around 15% for patients over 50 years, and 1% of these patients will present with critical ischemia.³⁵

The clinical evolution of patients with PAD rapidly increases with age and, therefore, an increase in the number of hospitalizations due to MLEA, especially elderly people.³⁰

The results of the incidence of MLEA observed in Rio de Janeiro correspond to 1/3 of the incidence in Sweden,³⁵ Denmark,³⁶ Finland,^37,38 and in the Unites States.³⁹ The results obtained in Rio de Janeiro correspond to half of the incidence in Taiwan,²⁸ with approximately 100/1,000,000/year, and around half of the incidence reported in diabetics in the United Kingdon.¹¹

Incident estimates of MLEA performed through the C-R technique, based on three data sources, for age group 30-89, showed rates of 102 and 21/100,000/year for DM and PAD, respectively. However, for age group 55-74 years, these estimates were 210 for DM and 360/100,000/year for PAD.

Concerning the general population of Rio de Janeiro (1992-1994), a study developed by Spichler et al.⁵ showed that MLEA incidence estimate per 100,000/year was 13.9 by the C-R technique and 5.4 by passive surveillance, when six etiologies were analyzed, PAD and DM, trauma, tumor, osteomyelitis and emphysematous gangrene. To complement this incidence analysis for the same period with age group 55-74, the estimate found was 31/100,000/year.

Calle-Pascual et al.,⁴⁰ using the same C-R technique, found incidences of 42 for men and 14 for women per 100,000 inhabitants/year, over forty years of age. For patients with PAD, these incidences were estimated in 1.1 and 0.6, for men and women, respectively. The author affirmed that such MLEA incidences on area 7 of Madrid are the lowest incidences reported in Europe, for both diabetics and PAD.

The results from our sample in Rio de Janeiro are similar regarding the general population, but they are 6 to 10 times higher in diabetics and PAD patients, respectively.

Using the C-R technique and based on two sources, an additional study estimated MLEA and showed that the highest rates were found among the Navajo Indian population in the United States, with 44/100,000/year. The study was conducted in 10 European, North American and Asian centers with population over 200,000 inhabitants. There are distinguished differences among the centers. It is important to emphasize the prevalence of PAD and diabetes mellitus among these populations.⁴¹

Incidence estimate from our sample, from the city of Rio de Janeiro. was lower than the incidence found among the Navajos in the Unites States, but it was higher than the incidences referred in other nine centers.

CONCLUSIONS

Although the prevalence of diabetes mellitus in the city of Rio de Janeiro is 9.2% for age group 30-89 years, major lower extremity amputations due to diabetes mellitus present under-reported diagnosis in the first five years, among our patients. MLEA had a significant increase, surpassing the major lower-limb amputations due to peripheral arterial disease, from 1995 onward, which represents 51.2%, that is, more than half of the last six years.

An early diagnosis and an adequate metabolic control of diabetes mellitus and critical ischemia, as well an increase of the number of primary or secondary revascularization surgical procedures, may postpone or even reduce the high number of primary amputations on the thigh in 70%. These procedures have an impact on the clinical evolution, mobilization, prosthetization, quality of living and reinsertion into society.

The incidence of major lower extremity amputations in the two etiologies occurs, predominantly, among males in all age groups and time periods, with a ratio of 1.5:1 and 2:1 for diabetes mellitus and peripheral arterial disease, respectively, which shows the deficiency of the health system regarding clinical follow-up for these etiologies.

There is a potential difference between the diagnosis and the adequate treatment of patients with critical ischemia and diabetes mellitus, which shows a five-time increase in the annual incidence of MLEA by passive surveillance (19 times in diabetes mellitus and three times in peripheral arterial disease) among age group 30-89 years.

Among age group 55-74, the incidence was three times higher than in the general population, and six times higher when diabetes mellitus was associated with peripheral arterial disease, which shows significant under-reporting of diabetes mellitus during the initial period of our study (1990 a 1994).

The C-R technique, when compared to passive surveillance, has found incidences 7 to 12 times higher, respectively, among diabetics and peripheral arterial disease population in the age group 55-74 than among the general population. The annual estimate of major lower extremity amputations was respectively: 31.3/100,000 inhabitants, 209/100,000 diabetics and 359/100,00 PAD, with a confidence interval probably closer to the reality than to passive surveillance, when compared to the general population. It could also be applied to epidemiological surveillance.

The incidence of MLEA should be considered as an important public health problem, requiring, therefore, complementary studies.

ACKNOWLEDGEMENTS

The authors would like to thank Dr. Yue Fang Chang of the Department of Epidemiology, University of Pittsburgh, PA, USA, for the statistical analysis of the capture-recapture technique.

REFERENCES

1. Weitz JI, Byrne J, Clagett GP, et al. Diagnosis and treatment of chronic arterial insufficiency of the lower extremities: a critical review. Circulation 1996;94:3026-49.

2. Lessa I. Introdução à epidemiologia das doenças cardiovasculares no Brasil. In: Lessa I. Epidemiologia das Doenças Crônicas Não-transmissíveis. São Paulo -Rio de Janeiro: Hucitec ABRASCO; 1998. p. 73-76.

3. Transatlantic Intersociety Consensus Revascularization - Management of the peripheral arterial disease. Eur J Endovasc Surg 2000;19 Suppl A:S1-250.

4. Malerbi DA, Franco LJ, and the Brazilian Cooperative Group on the Study of Diabetes Prevalence. Multicenter study of the prevalence of diabetes mellitus and impaired glucose tolerance in the urban Brazilian population aged 30-69 yr. Diabetes Care 1992;15:1509-16.

5. Spichler ERS, Spichler D, Lessa I, Forti AC, Franco LJ, LaPorte RE. Capture-recapture method to estimate lower extremity amputation rates in Rio de Janeiro, Brazil. Pan Am J Public Health 2001;10:334-40.

6. GLIM Statistical Package, 3.777. Update 2 (1985) London. Royal Statistic Society.

7. Hirsch AT, Criqui MH, Treat-Jacobson D, et al. Peripheral arterial disease detection, awareness, and treatment in primary care. JAMA 2001;286:1317-24.

8. McLafferty RB, Dunnignton GL, Mattos MA, et al. Factors affecting the diagnosis of peripheral vascular diseases before vascular surgery referral. J Vasc Surg 2000;31:870-9.

9. Ebaugh JL, Feinglass J, Pearce WH. The effect of hospital vascular operations capability on outcomes of lower extremity arterial bypass graft procedures. Surgery 2001;130:561-9.

10. Connelly J, Airey M, Chell S. Variations in clinical decision making is a partial explanations for geographical variations in lower extremity amputations rates. Br J Surg 2001;88:529-35.

11. The Vascular Surgical Society of Great Britain and Ireland. Critical limb ischaemia: management and outcome. Report of a national survey. Eur J Vasc Endovasc Surg 1995;10:108-13.

12. Cosgrove CM, Thornberry DJ, Wilkins DC, Ashley S. Surgical experience and supervision may influence the quality of lower limb amputation. Ann R Coll Surg Engl 2002;84:344-7.

13. Criqui MH, Browner D, Fronek MR, Coughlin SS, Barret-Conner E, Gabriel S. Peripheral arterial disease in large vessels is epidemiologically distinct from small vessel disease: an analysis of risk factors. Am J Epidemiol 1989;129:1110-9.

14. DHSS. Amputation Statistics for England, Wales and Northern Ireland. London: Department of Health and Social Security; 1986.

15. Jonason T, Ringqvist I. Factors of prognostic importance for subsequent rest pain in patient with intermittent claudication. Acta Med Scand 1985;218:27-33.

16. Spichler D, Spichler ERS, Lessa A, Forti AC, Franco LJ. Femoropopliteal Bypass in Critical Leg Ischemia Related Lower Extremity Arterial Disease with and without diabetes mellitus, in the City of Rio de Janeiro, Brazil, 1996-9. Diabetologia 2001;4 Suppl 1:I-IV A282-1083.

17. Widmer LK, Biland L, Da Silva A. Risk profile and occlusive periphery artery disease (OPAD). Proceedings of the 13th International Congress of Angiology; 1985 June; Athens, Greece.

18. Peabody CN, Kannel WB, McNamara PM. Intermittent claudication: surgical significance. Arch Surg 1974;109:693-7.

19. Farkouh ME, Rihal CS, Gersh BJ, et al. Influence of coronary heart disease on morbidity and mortality after lower extremity revascularization surgery: a population-based study in Olmsted County, Minesota (1970-1987). J Am Coll Cardiol 1994;24:1290-6.

20. Wolfe JN. Defining the outcome of critical ischaemia; a one year prospective study. Br J Surg 1986;73:321.

21. Verhaege R. Epidemiology and prognosis of peripheral obliterative arteriopathy. Drugs 1998; Suppl 3:1-10.

22. Cheng SW, Ting AC, Lau H, Wong J. Survival in patients with chronic lower extremity ischaemia: a risk factor analysis. Ann Vasc Sur 2000;14:158-65.

23. Catalano M. Epidemiology of critical limb ischaemia: North Italian data. Eur J Med 1993;2:11-14.

24. Toursarkissian B, Shiremann PK, Harrison A, D'Ayala M, Schoolfield J, Sikes MT. Major lower-extremity amputation: contemporary experience in a single Veterans Affairs institution. Am Surg 2002;68:606-10.

25. Faries PL, LoGerfo FW, Hook SC, et al. The impact of diabetes on arterial reconstruction for multilevel arterial occlusive disease. Am J Surg 2001;181:251-5.

26. Peel J, Stonebridge P. Association between age and survival following major amputation. The Scottish Vascular Audit Group. Eur J Vasc Endovasc Surg 1999;17:166-9.

27. Mayfield JA, Reiber GE, Maynard C, Czerniecki JM, Caps MT, Sangeorzan BJ. Trends in lower amputation in the Veterans Health Administration, 1989-1998. J Reahab Res Dev 2000;37:61-9.

28. Cheng SW, Ting AC, Lau H, Wong J. Epidemiology of atherosclerotic peripheral arterial occlusive disease in Hong Kong. World J Surg 1999;23:202-6.

29. McWhinnie DL, Gordon AC, Collin J, Gray DW, Morrison JD, et. al. Rehabilitation outcome 5 years after 100 lower limb amputations. Br J Surg 1994;81:1596-9.

30. Kihn RB, Warren R, Beebe GW. The "Geriatric" amputee. Ann Surg 1972;176:305-14.

31. Dormandy J, Belcher G, Broos P, et al. Prospective study of 713 below-knee amputations for ischaemia and the effect of a prostacyclin analogue on healing. Hawaii Study Group. Br J Surg 1994;81:33-7.

32. Spichler D, Spichler ERS, Martins CSF, Franco LJ, Lessa I. Diabetes mellitus and lower extremities amputation (1990-1996) in Rio de Janeiro, Brazil. Diabetologia 1998;41 Suppl1:I-IV A279.

33. Ebskov LB. The Danish amputation register 1972-1984. Prosth Orth Int 1986;10:40-2.

34. Fowkes FGR, Housley E, Cawood EH, Macintyre CC, Rucley CV, Prescott RJ. Edinburgh Artery Study: prevalence of asymptomatic and symptomatic peripheral arterial disease in the general population. Int J Epidemiol 1991;20:384-92.

35. Liedberg E. Age, diabetes and smoking in lower limb amputation for arterial occlusive disease. Acta Orthop Scand 1983;54:383-8.

36. Ebskov LB, Schroeder TV, Holstein PE. Epidemiology of leg amputation; the influence of vascular surgery. Br J Surg 1994;81:1600-3.

37. Pohjolainen T, Alaranta H. Lower limbs amputation in southern Finland 1984-1985. Prosth Orthot Int 1988;12:9-18.

38. Siitonen OI, Niskanen LK, Laakso M, Siitonen JT, Pyorala K. Lower extremity amputations in diabetic and nondiabetic patients. A population based study in eastern Finland. Diabetes Care 1993;16:16-20.

39. Centers for Disease Control: Prevalence, incidence of diabetes mellitus in the United States, 1980-1987. Morbid Mortal Wkly Rep 1990;39:809-12.

40. Calle-Pascual AL, Garcia-Torre N, Moraga I, et al. Epidemiology of nontraumatic lower-extremity amputation in area 7, Madrid, between 1989 and 1999. A population-based study. Diabetes Care 2001;24:1686-9.

41. Unwin N. Epidemiology of lower extremity amputation in centres in Europe, North America and East Asia. The global lower extremity amputation study group. Group TG. Br J Surg 2000;87:328-37.